In the study period from June 2016 to September 2020, a total of 262 cases underwent Total Knee arthroplasty. The majority were females, accounting for 74.4%, with ages ranging from 47 to 91 years old. Among the operated cases, 199 (76%) had at least two comorbidities, with hypertension being the most frequent, accounting for 67.6% of cases, followed by type 2 diabetes mellitus. Further details regarding the sample population are presented in Table 1. Regarding gender variations, both genders had similar baseline characteristics, as shown in Table 1. However, males tended to have lower comorbidity counts and lower BMI values. Nonetheless, 95% of the study population were overweight with a BMI over 25 kg/m2.

A surgical drain was applied in 156 patients (59.5%). The baseline characteristics of the patients, such as age, BMI, and comorbidities, were matched between the two groups, and no significant differences were found (P-value > 0.05). However, the group that received a drain had a significantly longer hospital stay (10.7 + -4.7 vs 5.4 + -4.6 days, P = 0.000), a 30% longer tourniquet time (111 vs 80 min, P = 0.000), a greater drop in haemoglobin (1.8 + -1.0 vs 1.2 + -0.7, g/dl P = 0.000) and haematocrit (5.4 + -3.0 vs 3.5 + -2.3,% P = 0.000), a higher average count of transfused packed RBC units (0.18 + -0.6 vs 0.04 + -0.2, units received P = 0.017), where each unit is of 250–300 ml per bag, and lower count of anticoagulants dose used (1.01 + -0.36 vs 1.13 + -0.50, units P = 0.025); i.e. daltaparin units. A detailed summary of the comparison between the drain and no-drain groups is presented in Table 2.

A sub-analysis based on tranexamic acid (TXA) use was conducted (Table 3). A total of 106 patients (40.6%) received TXA intraoperatively, based on surgeon’s preference and bleeding anticipation, and independent of drain use. Baseline characteristics such as age, BMI, and count of comorbidities were similar between the two groups (P-value > 0.05).

In the TXA group (n = 106), there was a significant reduction in hospital stays (6.3 ± 4.8 days vs 10.1 ± 5.1 days, p = 0.000, z = 6.6), shorter tourniquet time (88.2 ± 30.4 min vs 106 ± 24.2 min, p = 0.000, z = 4.6), less drain output (82.1 ± 167.9 mls vs 381 ± 263.7 mls, p = 0.000, z = 9.3), and higher levels of pre- and postoperative haemoglobin (1.4 ± 0.6 g/dl vs 1.6 ± 0.98 g/dl, p = 0.03, z = 2.2) and haematocrit (4.95 ± 3.1% vs 4.06 ± 2.1%, p = 0.016, z = 2.4) compared to the no TXA group (n = 156). Additionally, TXA use was associated with a higher absolute count of anticoagulation doses (1.21 ± 0.49 vs 0.96 ± 0.34, p = 0.000, z = -4.66). However, there was no statistically significant difference in the occurrence of postoperative complications between the two groups (P-value = 0.696).

20 cases had received blood transfusion (7.6%); with males 1.5 times more than females, as demonstrated in Table 2. Binary logistic regression demonstrated a statistically significant association between the following variables and the rate of blood transfusion: the BMI (Odds Ratio: 0.89, p-value = 0.015), the Pre-operative haemoglobin readings (OR: 1,p-value = 0.021), and the drain output (OR: 5.59, p-value = 0.03). This association is then modified when factoring for the use of tranexamic acid; among those with “no tranexamic acid”; the association of the BMI (OR: 0.90, p-value = 0.05), the pre-operative haemoglobin readings (OR: 0.61, p-value = 0.044) has persisted, while became insignificant for the Drain output (OR: 1, p-value = 0.104), however, in the group where “tranexamic acid” was used; the significance of all of them disappeared.